Centrosome amplification, the acquisition of three or more centrosomes, results in defective spindles, multipolar mitosis and aneuploidy. Aneuploidy generates a plethora of genetic alterations that culminates in cancer initiation and promotion. Centrosome amplification is present in most human tumors, including 80-90% human breast cancers and precursor mammary lesions. How that centrosome amplification arises is based on correlative evidence; for example, human breast cancers over-expressing Her2 (30% of all human cancers) have high frequencies of centrosome amplification. In addition, many altered gene products involved in human cancers result in centrosome amplification and chromosome instability in cell culture. Our published results showed that Ras induced centrosome amplification through the MAPK pathway in cultured cells. Our preliminary data showed that Ras, but not c-Myc induces centrosome amplification in precursor mammary lesions. Expression assays showed that human and mouse mammary epithelial cells expressing Her2, or Ras display ectopically expressed cyclin D1 and Nek2. Down-regulation of cyclinD1, Cdk4 and Nek-2 prevent Her2 and Ras- dependent centrosome amplification. We propose the following hypothesis: The Her2/Ras pathway trigger centrosome amplification and mammary tumor initiation by through a classical pathway involving Cdk4 phosphorylation of NPM, and through an alternative pathway involving ectopically-expressed Nek2. In specific aim 1 we will distinguish between a classical and an alternative pathway to centrosome amplification. In specific aim 2 we will assess how the alternative pathway impinges on centrosome amplification. Specific aim 3 will address how interfering with the classical pathway in vivo prevents centrosome amplification in pre-malignancy and suppresses mammary tumorigenesis.